An Approach to Forecasting Electricity Consumption at Railway Locomotive Depots

Energy management of locomotive depots is an urgent task for railway transport. The paper discusses the problem of forecasting data on electricity consumption by locomotive depots. The SARIMA model is used to solve this problem [1]. Autocorrelation function and partial autocorrelation function are used to find model parameters. The accuracy of the predicted calculations was assessed using the MAPE metric. The paper presents the results of computer modeling for energy consumption data of locomotive depots of one of the railways of the Russia. The novelty of the approach lies in the fact that for locomotive depots, energy consumption forecasts were designed without taking into account the specifics of production in locomotive depots. As a result of the work recommendations were proposed to reduce the consumption of electricity by using Metering Data Management System. The obtained technique can be applicable to other consumers of electricity in railway transport.